Introduction
The term metaverse (https://www.thefastmode.com/expert-opinion/28940-exploring-the-metaverse-s-infinite-possibilities-with-6g) refers to a virtual model of a physical system that enables interaction of various entities, such as virtual models (e.g., avatars) of mobile devices/humans, virtual models of static entities (e.g., smart homes), and interactive experience technologies (e.g., augmented reality). Such a virtual modeling in the metaverse provides many benefits (e.g., analysis and real-time resource management) for wireless systems (e.g., sixth-generation (6G) wireless systems) by effectively enabling the design trends of proactive learning and self-configuring wireless systems [1]. The design trends of proactive learning and self-configuring wireless systems are necessary to meet the diverse requirements of wireless system applications (e.g., brain-computer interaction, smart tourism, and industry 4.0) in terms of traditional quality of service (e.g., latency and reliability) and quality of experience (e.g., sense of physical experience) metrics [2]. A self-configuring wireless system refers to an efficient operation with minimum possible intervention from end-users/network operators. A self-configuring design can benefit from a metaverse virtual model by performing extensive experiments. On the other hand, proactive learning
In this work, the keyword “proactive learning” refers to learning meta space models before user requests.
is necessary to optimally utilize network resources (e.g., computing, communication, and energy resources) in response to highly dynamic environments and stringent latency requirements. To perform proactive learning, there is a need to train metaverse models before users request services. These pre-trained models can be obtained using a privacy-preserving machine learning scheme, namely, federated learning (FL). Next, these pre-trained models will be stored and used by the metaverse to serve end-users. Collectively, a metaverse will enable self-configuring design and proactive learning in 6G to enable emerging applications.